0:00

Fortune 500 companies trust Interoptic for

0:03

optical transceivers and cables. Since 2004,

0:05

Interoptic has provided high-performance optics and

0:07

cables at a fraction of the

0:09

cost of OEM gear. Interoptic products

0:12

are 100% tested and backed up

0:14

by real engineers. Work with the

0:16

optics experts at Interoptic. Find out

0:18

more at Interoptic.com/packet pushers. Welcome

0:21

to Heavy Networking the flagship podcast

0:23

from the packet pushers now in

0:25

our 15th year. If you've ever

0:27

argued about the optimal MTU size

0:29

to maximize tunnel throughput while avoiding

0:31

fragmentation. You found your tribe. I

0:33

am Ethan Banks and with me

0:35

is Drew Conry Murray, your hosts

0:37

for today's episode. You can connect

0:39

with us on LinkedIn or our

0:41

community slack group and please do.

0:43

On today's episode, we're chatting with

0:45

wireless ISP engineer Elijah Zeta. Elijah

0:47

had an interesting connectivity challenge to

0:49

solve for a remote mountain town and

0:52

not much budget to solve it with,

0:54

but he got it done by building

0:56

pretty much his own SDWAN using micro

0:58

tick boxes and zero-tier software. This is

1:00

the sort of solution you could build

1:02

yourself. And there's more to the story,

1:04

of course, and Elijah's going to tell

1:06

us all about it. So Elijah, welcome

1:08

to the show. Hey, man, this is

1:10

your first time in packet pushers on

1:12

heavy networking. Would you introduce yourself? Who are

1:14

you and what do you do? Well,

1:16

hey, Ethan, Drew. Thanks for having me

1:18

on. My name is, like I said,

1:20

Elijah, I'm a network engineer at a

1:22

wireless internet service provider called Airbridge Broadband

1:24

in North Central Central Idaho. We're

1:27

pretty remote and rural. We were about

1:29

an hour and a half to a

1:31

Walmart. So that's what I do. I

1:33

designed the towers and I started out

1:35

doing installs, infrastructure, building the

1:38

towers from the ground

1:40

up. The past few years I've been

1:42

doing the engine on the engineering

1:44

side of things. Yeah,

1:46

in Idaho for for those of

1:48

you that aren't familiar with United

1:51

States geography that is Northwest part

1:53

of America The northern tip would

1:55

be on the Canadian border and

1:58

it's like you said that rural

2:00

mountainous, that really describes the whole

2:02

state, is that fair? Yeah, and

2:04

in fact, we're in one of

2:06

those rural parts of the state.

2:09

There's one traffic light for the

2:11

whole state. Rumor is it was

2:13

put in for driving students. And

2:15

the county goes all the way

2:17

from the Oregon border all the

2:19

way to the Montana border. So

2:21

it's the largest county. I think

2:23

it's the largest geography county in

2:26

Idaho, and it's one of the

2:28

least populated. Wow. So can you

2:30

tell us a little bit more?

2:32

Obviously it's a small town, but

2:34

how does this town connect to

2:36

the rest of the world and

2:38

what are they relying on that

2:40

connectivity for? So previously before Airbridge

2:43

Broadman started, connectivity was mostly DSL

2:45

and there were a few other

2:47

wireless internet providers in the area

2:49

that were providing maybe like one

2:51

to. three megabits per second. Airbridge

2:53

broadband came in and started offering

2:55

anywhere between 10 to 50 megabits

2:57

per second. Now we're offering anywhere

3:00

between 100 to 300 by 300

3:02

megabits per second. Connectivity in this

3:04

part of the state is a

3:06

challenge because as of right now,

3:08

there is one fiber line running

3:10

into the area that's lumen. And

3:12

there have been times where a

3:14

farmer with a backhoe hit that

3:17

fiberline and it took out 911,

3:19

it took out us, it took

3:21

out Verizon, it took out everybody

3:23

because we're 100% single-threaded into the

3:25

area. We are working on the

3:27

state level and there's some certain

3:29

other projects that will bring redundancy

3:31

into the area, but that's still,

3:33

we're probably still multiple years out

3:36

on that. Does the rural broadband

3:38

initiative at the federal level help

3:40

maybe with that? It does. A

3:42

lot of the projects that will

3:44

bring fiber redundancy into the area

3:46

have already been funded through previous

3:48

grants and previous years. Bede funding

3:51

that's coming down the pike is

3:53

mostly meant for last mile to

3:55

the home connectivity. So we do

3:57

have a few. projects that have

3:59

been thunded already, but as all

4:01

that goes, it still takes multiple

4:03

of years to, you know, engineering

4:05

studies, digging, actually getting in the

4:08

ground, and then operational, it's gonna

4:10

be a long time. Now you

4:12

said that you got one fiber

4:14

going into the area. What is

4:16

the, what is the use case

4:18

for the fiber? What kind of

4:20

traffic's going over that thing? The

4:23

majority of the traffic is mostly

4:25

like Lumen is using it to

4:28

connect their COs together. So Lumen

4:30

uses that fiber to go from

4:32

CO to CO. There are very

4:34

few organizations that have actual fiber

4:36

connectivity from Lumen's CO to whatever.

4:39

So we obviously have fiber connectivity

4:41

to a few locations. that we

4:43

then bounce wirelessly out. You might

4:45

have a few other locations that

4:47

have fiber, but for the most

4:50

part, it's mostly just looming CEO

4:52

to CEO. And then your users,

4:54

the people that are your customers,

4:56

are these business, home, both? Yep,

4:58

we serve residential business. It doesn't

5:01

really matter. Same wireless plans for

5:03

everybody. We don't make a difference

5:05

between wireless or residential. And how

5:07

do they access your service? So

5:09

they really just call us up

5:12

and say, hey, we want better

5:14

or faster internet. And then we

5:16

say, sure. So we go out

5:18

to their place, we see if

5:20

they can see a tower. So

5:23

we'll put a tower on a

5:25

hill or some other location that

5:27

can see a large amount of

5:29

users. We also prefer to do

5:31

a lot of little towers instead

5:34

of one monstrosity. So the majority

5:36

of our towers are around 40-ish

5:38

feet tall. We have one that's

5:40

80 and we have one that's

5:42

90. but the majority of them

5:45

around 40 feet and we prefer

5:47

to just do a lot of

5:49

little ones versus one or two

5:51

big ones and so we'll just

5:53

put a wireless fixed wireless antenna

5:56

on the building and shoot it

5:58

to the closest tower. So

6:02

in this area, you guys are

6:04

rural, spread out, mountainous region, you're

6:06

just trying to provide internet access

6:08

for everybody for all the reasons

6:10

that we all want internet access,

6:13

and then winter comes, and what

6:15

are the challenges? Yeah, so the

6:17

challenges are pretty good. We get

6:19

a lot of snow, ice in

6:21

this part of the country, and...

6:23

Well, to find a lot. Okay,

6:25

yeah, so where I'm at here,

6:27

I'm at around 2,200 feet, I

6:30

believe. I can get up to

6:32

a foot, maybe a bit more

6:34

of snow in the winter. Grangeville,

6:36

which is the town that Airbridge

6:38

Broadband is headquartered. We can get,

6:40

they can also get a little

6:42

bit over a foot. And then

6:44

we have a town called Elk

6:46

City, that's about an hour and

6:49

a half from Grangeville. It takes

6:51

about a 25 mile point-to-point link

6:53

from... Grangeville to a top of

6:55

a mountain and then another I

6:57

can't remember what it is it's

6:59

like another sevenish miles from the

7:01

top of a mountain down to

7:03

Elle City and they can get

7:06

a couple feet of snow and

7:08

they're about an hour and a

7:10

half away from Grangeville and so

7:12

so the only way into that

7:14

town is the the point-to-point shots

7:16

over the mountain ranges yeah and

7:18

in fact that's even how the

7:20

incumbent gets there like their their

7:23

DSL and their cell phone service

7:25

into that town. That's how the

7:27

incumbent does it. They shoot a

7:29

wires link up to the top

7:31

of that mountain and then back

7:33

down as well. They're the only

7:35

physical lines running to the town

7:37

are power. There's no, you know,

7:40

the incumbent gets down into the

7:42

town and then they run, you

7:44

know, copper around the town, but

7:46

it's all wireless into the town.

7:48

Okay, so you so so foot

7:50

of snow, two feet of snow,

7:52

you mean at a time or

7:54

like through the whole season? That's

7:56

that's about what you're going to

7:59

get. Usually through the whole

8:01

season, sometimes more, sometimes less. And

8:03

it depends on the area, of

8:05

course, but yeah, that's usually throughout

8:07

the whole season. Okay, you're not

8:10

getting crushed by snow like some

8:12

of the, some of the West

8:14

Coast towns or Utah or Colorado

8:16

can have many, many feet of

8:18

snow. It's not that bad, but

8:20

it's still a, it's still a

8:22

thing. Yeah, it's still thinking it

8:25

can get really cold here, so

8:27

part of the. What's really weird

8:29

about this location too is we

8:31

can get some really extremes. We

8:33

can have negative 15 degree weather

8:35

in the winter and we can

8:38

have 120 degree weather in the

8:40

summer. And so the extremes and

8:42

temperature are equipment. We have to

8:44

make sure our equipment can work

8:46

with the extreme temperature differences, which

8:48

is one of the reasons we

8:50

chose micro tick as our main

8:53

routing and solutions platform is because

8:55

their equipment can handle. those extreme

8:57

temperature differences. Okay, so micro-tick boxes

8:59

end up, where are you using

9:01

them? Like you were talking about

9:03

wireless shot going from the top

9:06

of a mountain down into the

9:08

towns, you have micro-tick boxes on

9:10

either end of that? Yes, yeah,

9:12

so every tower gets a micro-tick

9:14

router and most of the time

9:16

a micro-tick switch as well. So

9:18

we're using it for routing and

9:21

switching, so micro-tick boxes at every

9:23

tower location. Most of the wireless

9:25

is ubiquity. Even that 25 mile

9:27

shot is a ubiquity air fiber

9:29

11 with a four foot dish

9:31

on each side. Oh, okay. Yeah,

9:33

yeah. And so you said obviously

9:36

being able to handle extreme temperatures

9:38

is one of the reasons you

9:40

went with a micro tick, but

9:42

I'm sure any of the major

9:44

brand name vendors would also be

9:46

happy to sell you hardened gear.

9:49

What else drew you to micro

9:51

tick? The cost. Of course, the

9:53

cost is a big thing with

9:55

my critique as well. It's a

9:57

great value for value for your

9:59

for the money. And it works

10:01

really well. The devices are hardened

10:04

that provides the feature set that

10:06

we need and they can push

10:08

the amount of traffic that we

10:10

need. And when configured correctly, they

10:12

work really well. They're super stable.

10:14

We've had micro-tech routers and switches

10:17

out there for years with over

10:19

a year of uptime because we'll

10:21

have battery backups at our tower

10:23

sites. And most of the time,

10:25

this stuff just gets rebooted mostly

10:27

due to a firmware update. you

10:29

know, every year or two. Let's

10:32

pause for a message from our

10:34

sponsor Interoptic. Interoptic is the optical

10:36

transceiver and cable specialist that maximizes

10:38

your IT savings while minimizing network

10:40

failures. Interoptic provides high performing optics

10:42

at a fraction of the price

10:45

of brand name optics. The Interoptic

10:47

experts can help you spec the

10:49

best optical transceivers and cables for

10:51

your network environment. Interoptic optical transceivers

10:53

are 100% guaranteed to be operationally

10:55

equivalent with Cisco, Juniper, Extreme, Arista

10:57

Brocade, Palo Alto, and many other

11:00

switches and routers. Due to Interoptic's

11:02

deep optical transceiver technical experience, they

11:04

can ensure that all messages, alerts,

11:06

alarms, alarms, and threshold data are

11:08

equivalent to OEM brands. Interoptic deploys

11:10

rigorous, 100% testing on their devices

11:12

before they're shipped. Interoptic optical transceivers

11:15

are built to the exact same

11:17

manufacturing lines. insurance companies, retailers, financial

11:19

services, and federal and state government

11:21

customers deploy optics and cables from

11:23

Interoptic. You can purchase the same

11:25

if not better performing optical transceivers

11:28

tested and backed up by real

11:30

engineers at a fraction of OEM

11:32

costs. Find out more at interoptic.com/packet

11:34

pushers. That's interoptic.com/packet pushers. And now

11:36

back to the podcast. Now, Elijah,

11:38

when you chatted with us about

11:40

this solution that you built this,

11:43

because we haven't talked about the

11:45

zero-tier component yet, we're going to

11:47

get there, but zero-tier and micro-tech

11:49

has an integration, and you use

11:51

that to help out elk city,

11:53

as I remember right in the

11:56

winter. This was the city that

11:58

we're getting to via. the wireless

12:00

shot and that's it, because the

12:02

only hard lines going into the

12:04

town are power, as you described.

12:06

So paint the scenario for us

12:08

where Elk City might go off

12:11

the air, have some trouble because

12:13

of the weather. Yeah, so we have

12:15

the 25 mile shot going from

12:17

Grangeville to the top of this

12:20

mountain, another 5-7-ish miles down into

12:22

town. Well, this mountain, the top

12:25

of this mountain is also strictly

12:27

and only powered by solar. Oh

12:29

wow. And the only way to get up

12:31

to the top of the mountain,

12:34

probably a good six to seven

12:36

months out of the year, is

12:38

either snowmobiles or helicopter.

12:40

And in fact, so we're renting

12:42

space on a tower up there

12:45

and they have a generator up

12:47

there, but almost every year, the

12:49

generator either runs out of propane. Last

12:51

year, the generator burnt up, went

12:53

through all of its oil and

12:55

they had to helicopter in more

12:57

oil. to the top of the

13:00

mountain to get the generator

13:02

up and running. But you know, when

13:04

your solar panels have two, well up

13:06

there, it's probably five or six feet

13:08

of snow up on top of this

13:11

mountain, the panels just don't charge and

13:13

you have to run the generator and

13:15

the generator goes down, it was

13:17

all of your equipment. And so

13:20

the one link into this town

13:22

goes out and everybody loses their,

13:24

is their connectivity. And so,

13:26

as you mentioned, we came

13:28

up with a solution called

13:31

Zero Tier and Zero Tier

13:33

is an SDWAN VPN. They

13:35

call it like a smart

13:37

switch for planet Earth is

13:40

what they call it. And

13:42

it's essentially a layer two.

13:44

solution that we can put it

13:46

has a micro tick has a

13:48

native package for zero tier that

13:50

you can install on these boxes and

13:53

it essentially makes it as if

13:55

it's a physical interface on both of

13:57

the boxes it's end-to-end encrypted using

13:59

cryptographic ID and I can make it

14:01

look as if the two are

14:03

connected with a physical interface and

14:05

I so then I can have

14:07

a micro tick box at our

14:10

head end we have a micro

14:12

tick box in elk city and

14:14

so then using something we'll

14:16

say star link for an

14:18

example I can have a

14:20

star link dish at our

14:22

main site and when the

14:24

main connectivity over this mountain

14:26

goes offline well The zero

14:29

tier will automatically reroute

14:31

through the Starlink and

14:33

I can run OSPF,

14:35

IPGP, whatever I want

14:38

over the Starlink interface

14:40

and then which keeps

14:42

the customers online until

14:44

the tower gets back

14:46

up. So it's a very

14:48

effective solution. We use zero

14:50

tier for a VPN type

14:53

as well. So I use it to,

14:55

I have a zero tier app on

14:57

my phone. and I can use it

14:59

to VPN in our network so that

15:01

I can be able to administer

15:04

our network from really anywhere

15:06

in the world. And what's great

15:08

is it's kind of like

15:10

I trippily 802.1x I believe

15:12

it is, and where it

15:14

has like port-based network access

15:16

controls where, you know, most VPNs.

15:18

you kind of just give a person

15:21

a username and password and then they

15:23

you know who knows like who they

15:25

could give that user even password out

15:27

to they could use that username and

15:30

password on their phone their computer their

15:32

friend who knows right zero tier is

15:34

great because each device has its

15:36

own authentication and so I can

15:38

give somebody the network ID and

15:40

they can put it on their phone

15:42

but if they want to take that

15:45

network ID and join on their computer

15:47

it wouldn't work because I haven't authorized

15:49

that device. And so, but anyway, so

15:51

that's a quick introduction on that and

15:54

it's works really well. And in fact,

15:56

just really quick, we had this similar

15:58

incident the other day. where we had

16:00

a wildfire and a tower side of ours

16:03

and took the tower offline because the

16:05

power company shut all the power down

16:07

and a similar situation we set up

16:09

a star link and got zero tier

16:11

up and running and we were able to

16:13

provide internet to the town for the

16:16

short while while the power company got

16:18

the power back online. Zero tier just

16:20

to put it's basically it's an

16:22

overlay it's a tunneling solution and

16:25

it tunnels Ethernet frames in their

16:27

entirety is that right? That's

16:29

correct. It is a it is a layer two

16:31

tunnel. And so what's great about

16:33

that is being a layer two tunnel,

16:36

like micro take has what they

16:38

call an MDP packet micro take

16:40

discovery protocol, which is very similar

16:42

to CDP and all those other

16:45

discovery protocols. And so because it's

16:47

a pure layer two tunnel, and

16:49

it uses UDP and it can

16:52

punch through just about anything DNS,

16:54

double triple that. Obviously, what's great

16:56

about it is if you can

16:58

open up your UDP ports, and

17:01

if you aren't doing like multiple

17:03

NAT, then it will do a

17:05

point-to-point connectivity. So it'll like connect

17:07

itself using their with their, they

17:09

call them planets and moons. So

17:11

like their servers, the two endpoints

17:13

will connect through their servers. But

17:16

if you can open up your

17:18

UDP ports, they will provide direct

17:20

point-to-point connectivity. It won't even go

17:22

through their servers at all. I

17:24

see what you're saying. So there's

17:26

two different architectures then with zero

17:28

tier. You can hit one of

17:30

their planets. It sounds like a

17:32

hub that acts as a relay

17:34

to connect your two endpoints. Or

17:36

as you're saying, if you've got UDP

17:38

connectivity available end to end, then you're

17:40

standing up a point to point tunnel

17:43

direct without having to go through any

17:45

of the zero tier. I'll call it

17:47

the cloud architecture. Exactly, and

17:49

you can even spin up your

17:51

own cloud architecture if you want.

17:54

You don't have to use theirs.

17:56

But like I said, it's still

17:58

all end-to-end encrypted using unique talk.

18:00

cryptographic ideas and everything like that.

18:02

So I just let them handle

18:05

the cloud infrastructure. Okay. Now, micro

18:07

tick is inexpensive gear. As you

18:09

said, we've had, we've had some

18:11

other wireless ISP guys, come on

18:13

and talk about micro tick before.

18:16

And that was, that was another

18:18

thing that they cited for sure

18:20

as you did. Cost. You can

18:22

get a lot of functionality, a

18:24

lot of bang for the buck

18:27

with micro tick. How does zero

18:29

tier fit into that? If I

18:31

remember, back in the day, it

18:33

was open source with a commercial

18:35

variance, something like that, something like

18:38

that, something like that. Yeah, so

18:40

the way zero-tier integrates with it

18:42

is micro-take came out with a

18:44

native zero-tier package. So I can

18:47

install zero-tier directly on the micro-tick

18:49

box. Now, zero-tier does have, you

18:51

know, they have iOS, Android apps,

18:53

Windows apps, Linux apps, Mac, you

18:55

can run it on a synology

18:58

and docker, BSD. I mean, it's

19:00

very versatile. I know a guy

19:02

who's running it like on a

19:04

prox-aox. box because zero tier runs

19:06

on Linux natively too. So he

19:09

has some like ubiquity edge routers

19:11

and he can he'll he'll go

19:13

into the CLI into the ubiquity

19:15

edge router because it's just dabbing

19:17

in the background and he'll run

19:20

the zero tier Linux script CLI

19:22

of his edge router and he'll

19:24

get zero two up and running

19:26

on his ubiquity edge router. And

19:28

so it's very versatile if you

19:31

can get to the CLI. I

19:33

mean, I think he even talked

19:35

about getting it running on like

19:37

an ERISA switch if you can

19:40

get down into the CLI. And

19:42

so, you know, ProxMox, you know,

19:44

you get zero tier working on

19:46

a ProxMox device and you now

19:48

have a zero tier connectivity between

19:51

ProxMox and whatever else. It just

19:53

shows up as a physical interface.

19:55

And you can. literally run whatever

19:57

you want on top of it.

19:59

OSPF, IBCP, E, MPLS, VPLS, whatever,

20:02

whatever you like. So I want

20:04

to make sure I understand how

20:06

this is working if there's a

20:08

tower that's providing service to Elk

20:10

City and the tower power goes

20:13

out for whatever reason out of

20:15

this generator snow whatever fire How

20:17

is zero tier helping because if

20:19

there's no power to what is

20:21

zero tier doing or how are

20:24

you overcoming the power issue? So

20:26

we go from the top of

20:28

the mountain to one tower down

20:30

into the town and then from

20:33

that one tower in the town,

20:35

we go to about four or

20:37

five other towers. And so all

20:39

four or five other towers in

20:41

the town go to one tower,

20:44

that one tower goes to the

20:46

top of the mountain. The one

20:48

tower that's the, kind of, we'll

20:50

say it's the hub down in

20:52

the city is a starling dish.

20:55

goes to a micro tick router

20:57

which has that on it that

20:59

provides so that that's how we

21:01

can get internet to that micro

21:03

tick box through that star link.

21:06

Well you put I put a

21:08

zero tier package which then doesn't

21:10

zero tier interface and using zero

21:12

tier IDs and everything you kind

21:14

of connect it to another micro

21:17

tick box on the another part

21:19

of our network. Well I just

21:21

put like we'll say OSPF on

21:23

those zero tier interfaces. Well now.

21:26

the microtrick box in Elksay on

21:28

that tower is connecting to another

21:30

microtrick box on another part of

21:32

our network using OSPF over zero

21:34

tier over Starlink and so and

21:37

because of like you know what

21:39

we'll say OSPF right has a

21:41

cost associated to it so a

21:43

higher cost value it won't the

21:45

traffic you know it's the traffic

21:48

won't go over that so the

21:50

OSPF cost on that link is

21:52

really really high so then So

21:54

traffic will preferentially prefer the main

21:56

link going over the tower and

21:59

over the the mountain. Well, then

22:01

if that ever goes down, OSPF

22:03

will. reroute the traffic. Actually, it's

22:05

just the management traffic. Really, the

22:07

customer traffic is netted through the

22:10

Starlink itself, and it's just management

22:12

traffic that will go over OSPF,

22:14

over the Starlink, and over so

22:16

we can still maintain management access

22:19

to all of our devices for

22:21

the day or so that hours

22:23

out. So

22:25

rather than bouncing it up across

22:27

the mountain via the tower, you're

22:29

bouncing it up to lower-th-orbiting satellite

22:31

and back down because you've got

22:34

two satellite, two star-link base stations,

22:36

one close by and one up

22:38

in Elk City? Well, yeah, the

22:40

one that's close by, actually, we

22:42

don't have one close by. The

22:44

microtech box is just mostly at

22:46

our head end, and so it's

22:48

just fiber connectivity. There's no star

22:50

link there. The star link is

22:53

only on the Elk City side.

22:55

Yeah, at some point it gets

22:57

into the Starlink network though, and

22:59

gets, and then gets beamed down

23:01

to them. So they're just, they're

23:03

just another, it just so happens

23:05

the last mile is Starlink for

23:07

Elk City in that case. That's

23:10

their backup. Yep, for the, for

23:12

the day or two, the power

23:14

goes out, we'll have Starlink there

23:16

and it works really well. So

23:18

I think we have like the

23:20

business package, you know, there that.

23:22

Okay, and you know, for the,

23:24

it works, it only works for

23:26

a day or two at the

23:29

year. It just kind of sits

23:31

there pretty much all of the

23:33

rest of the months. But for

23:35

the day or two, the year

23:37

that we needed, it works really

23:39

well. Yeah, do you feel like

23:41

at some point you could use

23:43

that as an additional link to,

23:46

to provide more traffic, like, as

23:48

opposed to just a backup or

23:50

an emergency option? Maybe we'd have

23:52

to kind of we'd have to

23:54

look I think it's best mostly

23:56

only use it as a backup

23:58

or emergency solution It's best to

24:00

probably use our own infrastructure and

24:02

have the traffic over our own

24:05

devices primarily And so we'll probably

24:07

just keep it as a backup

24:09

solution Okay And just so I'm

24:11

understanding the architecture, the Starlink dish

24:13

is not on the tower that

24:15

has the energy problems. It's in

24:17

the town. No, okay. No, no,

24:19

this Starlink is on a tower

24:22

that's actually in the town. It's

24:24

kind of like the main tower

24:26

hub in town. Okay. And then

24:28

from there, it goes like seven

24:30

miles up to the top of

24:32

the mountain. And then from the

24:34

top of the mountain is where

24:36

the energy issues are, because then

24:38

you have another 25 mile shot

24:41

down to. down to Grangeville. Okay.

24:43

And as you said, you're able

24:45

to use OSPF to set this

24:47

up so that it's, it's not

24:49

an always on path that only

24:51

comes on as needed. Exactly. Good.

24:53

Yeah. Yeah, latency. It feels like

24:55

both of these solutions could have

24:57

some significant latency, whether you're going

25:00

over the mountain with the two

25:02

different wireless shots or whether you're

25:04

using Starlink. So how is that

25:06

experience? Lateency is great when using

25:08

our infrastructure actually, you know, this

25:10

probably maybe a one to two

25:12

millisecond shot, the 25 mile shot.

25:14

It's really not that bad. You

25:17

know, the seven mile shot is

25:19

going to be under a millisecond.

25:21

And so when using our own

25:23

infrastructure, latency is great. And when

25:25

using the zero tier and Starlink

25:27

solution, it's one of the reasons

25:29

I opted to actually, so the

25:31

the microtrick box that has the

25:33

zero, the microtrick box. that has

25:36

the star link attached to it

25:38

actually has some net rules that

25:40

will not customer IP traffic through

25:42

the star link. And so it's

25:44

one of the reasons I opted

25:46

to do that and then only

25:48

have management access go over OSPF

25:50

is to help is to help

25:53

with latency because otherwise customer traffic

25:55

is going to flow over OSPF

25:57

star link and then have to

25:59

route all the way back into

26:01

our network and then route all

26:03

the way back out versus just

26:05

that. through the Starlink dish and

26:07

then management access going to be

26:09

a bit slower and have latency.

26:12

but I'll deal with that for

26:14

the day or two. And so

26:16

that's how I dealt with the

26:18

lane C problem. Just minimizing the

26:20

amount of traffic that, okay, so

26:22

in that scenario, what you're saying

26:24

is the elk city, our population

26:26

here, they're gonna get their internet

26:29

access via Star Lake direct, not

26:31

added, but if it's coming back

26:33

to you as the provider, it's

26:35

again, all you're really concerned about

26:37

is the management traffic and keeping.

26:39

keeping the boxes up and monitored

26:41

and so on. Exactly. Yep. Just

26:43

so I can monitor all the

26:45

boxes, if I have to log

26:48

into a switch, access point, you

26:50

know, like troubleshoot customer issues, you

26:52

know, all of our tech teams

26:54

can still have access to all

26:56

of the devices. So what you

26:58

don't have in this solution is

27:00

like application based routing. Like if

27:02

you, if this is a full

27:04

blown SDWAN solution, that would be

27:07

one of the things. Oh, I

27:09

want to send. Web traffic over

27:11

Starlink because it's fine, but anything

27:13

that's more latency sensitive I'm gonna

27:15

push it this other direction Over

27:17

the tower like if both solutions

27:19

were up. I'm just I think

27:21

for like how could you have

27:24

both solutions up and working? You

27:26

know, would you want you could

27:28

if you wanted a zero tier

27:30

has a lot of like policy

27:32

based? Firewall type rules and things

27:34

like that. And then a micro

27:36

ticket course has a lot, they

27:38

call them like mangle rules and

27:40

layer seven application firewall rules. You

27:43

could do just about anything you

27:45

wanted. Your your imagination is probably

27:47

the limit to what you can

27:49

do with zero tier and micro

27:51

tick. But for me, it's a

27:53

little bit simpler. Well, I consider

27:55

it a little bit simpler. I

27:57

would just rather want all traffic

28:00

to go over the mountain when

28:02

it's up, if it's down. I

28:04

want customer traffic to go over

28:06

the star link and then management

28:08

access to go over zero tier

28:10

OSPF and that. As soon as

28:12

you start introducing some kind of

28:14

a manual policy base. routing where

28:16

you're picking this source and that

28:19

source to go through this line

28:21

or that line and then you

28:23

got to troubleshoot it because reasons.

28:25

Yeah, no, you don't want to

28:27

do that if you don't have

28:29

to, but not really good compelling

28:31

reason to do that. You'd need

28:33

something. You'd need a, again, like

28:36

a full blown SDWAN solution where

28:38

you could build that policy at

28:40

a higher level and keep everybody

28:42

saying if something's going wrong, trying

28:44

to troubleshoot it. You zero tier

28:46

can install as a package on

28:48

micro tech. Does that mean it's

28:50

a, it sounds like it's a

28:52

micro tech supported solution. They support

28:55

this thing natively. Yeah, they build

28:57

the package. You just literally drag

28:59

and drop it over to the

29:01

router, you reboot the router, and

29:03

you now have the zero tier

29:05

native package on micro tech. Yes,

29:07

they support it. And that's one

29:09

of the reasons micro tech. I

29:12

love microtrick and it's not for

29:14

everybody, but it's the right tool

29:16

for us. And one thing that

29:18

microtrick doesn't have is a phone

29:20

number that I can call and

29:22

say, hey, I need something, right?

29:24

And but that's one of the

29:26

reasons they're able to keep their

29:28

costs slow. You know, call center

29:31

support costs a lot and they're

29:33

able to keep a lot of

29:35

their costs slow by not having

29:37

that. So you have forums and

29:39

you have the knowledge base and

29:41

everything. So if you know how

29:43

to use it's. really powerful you

29:45

can get a lot done with

29:47

it and yeah they just have

29:50

a native package that I install

29:52

and it works great. So do

29:54

you have to pay licensing cost

29:56

to zero tier for your use

29:58

case? Zero tier they changed their

30:00

licensing or price and structure up

30:02

recently so I'm not 100% familiar

30:04

with it as I used to

30:07

be but before you had to

30:09

pay like $5 a month if

30:11

you wanted to have multiple different

30:13

administrators or something like that. And

30:15

like I said, they changed their

30:17

pricing up recently, but it's really

30:19

price effective. I think we pay,

30:21

gosh, like 15 bucks a month

30:23

and that handles all, you know,

30:26

our zero tier through. So we

30:28

have our backup network, we have

30:30

our VPN network. use zero tier.

30:32

So I have a micro, so

30:34

Airbridge broadband has three separate offices,

30:36

one in Kami, I, Lewiston, Grangeville,

30:38

and each office has a micro

30:40

tick router for the main router,

30:43

and I have a zero tier

30:45

package on each office, and zero

30:47

tier, I have connected all of

30:49

the offices together. So I can

30:51

print on a printer in Lewiston,

30:53

if I'm in Grangeville, I can

30:55

access the internal network. All the,

30:57

all three offices are connected through

30:59

zero tier. What about support for

31:02

zero tier that they also have

31:04

similar situation? Yeah, they have their

31:06

documentation. They have a great API

31:08

community and things like that. Honestly,

31:10

I've never really needed to have

31:12

a reason to try to even

31:14

try to reach out to them.

31:16

Everything is just super simple. You

31:19

basically log in at the Create

31:21

a network button. They give you

31:23

a network ID. you have the

31:25

app on your phone, your computer,

31:27

micro check, whatever, you copy and

31:29

paste in that network ID, you

31:31

log into the website, you hit

31:33

the authorized button to authorize the

31:35

device, and kind of off to

31:38

the race. So they also have,

31:40

the one thing that was a

31:42

little bit tricker to figure out,

31:44

but even that wasn't that bad,

31:46

they have what are called managed

31:48

routes, and oh, it can access

31:50

as a like even a DHCP

31:52

server. So like the VPN solution

31:54

that we have, you know, you

31:57

go in there and you tell

31:59

it, hey, I want to be

32:01

able, I want, I want you

32:03

to be able to access, we'll

32:05

take the, the 172 range, right,

32:07

the private IP range of 72.

32:09

So, you know, 172, 16,00 slash

32:11

12, I want to be able

32:14

to access that and you, you

32:16

route any of that traffic through

32:18

a specific IP. So like, we

32:20

have a, are called a VPN

32:22

router. has a specific IP address

32:24

and then I say I want

32:26

to be able to route 1716.00

32:28

slash 12 through this IP and

32:30

so then on my phone, computer,

32:33

whatever, I type in 172, 16,

32:35

whatever, whatever, whatever, then it knows

32:37

to route that traffic specifically through

32:39

zero-tier, through that other IP, and

32:41

then obviously the routers know that

32:43

they can access that 172 range

32:45

through that IP address. Now you

32:47

mentioned zero-tier is essentially doing layer

32:50

two encapsulation, so it's tunneling. Do

32:52

you ever run into concerns with

32:54

tunneling? Does it affect performance ever?

32:56

Does it cause troubleshooting headaches? Does

32:58

it cause of routing headaches? Nope,

33:00

never had any problems with it

33:02

because it's a it's a layer

33:04

two tunnel. And so if I

33:06

remember right, they're doing like jumbo

33:09

frames. You can do just about

33:11

anything you want with it and

33:13

you never have a problem. You

33:15

can just put anything you want

33:17

for it. Yeah, that's what I

33:19

was going to ask you about

33:21

with fragmentation. Do you have to?

33:23

deal with MTU and MSS sizes

33:26

on interfaces, anything like that to

33:28

make sure that you're going to

33:30

be able to get that zero

33:32

tier frame or packet through unfragmented.

33:34

I've never had to deal with

33:36

fragmentation issues. I've had to done

33:38

this VPM thing, you know, traffic

33:40

and a bunch of stuff with

33:42

it. I've never had to deal

33:45

with fragmentation issues, but it just

33:47

depends on what you're doing. So

33:49

like you've talked up a lot

33:51

about zero tier and all the

33:53

cool stuff that you can do

33:55

with it. If I'm a network

33:57

engineer and I've not worked with

33:59

zero tier before, do you have

34:01

any tips or other things I

34:04

should be looking at or considering

34:06

as a network engineer to make

34:08

my zero tier deployment smooth? So.

34:10

Being as it's so simple to

34:12

get up and running, there's actually

34:14

not a whole lot. You basically

34:16

just need to go to zero

34:18

tier.com, create an account, you can

34:21

log in, create a network, and

34:23

then I think the biggest thing

34:25

is just making sure your managed

34:27

routes are set up correctly and

34:29

your users are set up correctly.

34:31

And you can run to their

34:33

documentation. They've got great documentation and

34:35

then documentation for whatever platform you're

34:37

using, like say MicroTEC. MicroTEC has

34:40

a great zero-tier page and you

34:42

can. It's super, like a microtrick,

34:44

it's super easy to install, it's

34:46

just any other package. If you're

34:48

used to microtrick, you just drag

34:50

and drop the package into the

34:52

folder location, reboot the router, and

34:54

it automatically installs the package. And

34:57

you basically take that zero-tier ID,

34:59

put it into the router, and

35:01

it'll connect to your zero-tier network,

35:03

and you just make sure your

35:05

managed routes are set up correctly,

35:07

and it's pretty flawless. Okay, so

35:09

zero tier ID. So is that

35:11

my endpoint identifier? Is that like

35:13

the network that I'm participating in?

35:16

That's like you're the network that

35:18

you're participating in. And so you

35:20

have a micro tick on one

35:22

end and a micro tick on

35:24

another, you're going to take that

35:26

network ID and you're going to

35:28

put it into both of them.

35:30

And so then. It tells both

35:33

those microticks that they are participating

35:35

in the same, we'll say, we'll

35:37

call it network in the same

35:39

network. And you know, I have

35:41

like, you know, again, I've like

35:43

our Grangeville office, Camie I office,

35:45

and Luston office, they're all participating

35:47

in the same network, so to

35:49

speak. Like I go to the

35:52

neighbors, I see all the offices.

35:54

it from one device. And so,

35:56

or like our VPN solution that

35:58

we have, so everybody that I

36:00

had given access to the VPN,

36:02

I give the same network ID

36:04

out to all of them. But

36:06

each device is authorized on a

36:09

per device basis. Got it. Okay.

36:11

So that means I could, if

36:13

I want to run more than

36:15

one zero-tier environment, more than one

36:17

zero-tier network on my network, I

36:19

could have just different IDs and

36:21

they would be unique from one

36:23

another. Exactly. Yeah, you just go

36:25

into your zero-tier account and you

36:28

just click on the Create a

36:30

network button and hop up another

36:32

network and different with a different

36:34

ID and... You can do whatever

36:36

you want. And then like our,

36:38

again, our VPN solution, I have

36:40

it set up as like a

36:42

DHCP server as well. So when

36:44

a device connects to the zero-tier

36:47

network, zero-tier will hand out an

36:49

IP address to that device. And

36:51

with that IP address, you know,

36:53

that IP is going to be

36:55

authorized to log into the rest

36:57

of the network. So could you

36:59

use that capability to, you know,

37:01

either do some kind of micro

37:04

segmentation or set up different kinds

37:06

of service tiers that you could

37:08

then offer to customers, you know,

37:10

like has, there's our highest service

37:12

tiers, slightly lower, etc. You probably

37:14

couldn't do that, but what you

37:16

could do, you could offer maybe

37:18

different services to the customer. So

37:20

like if you put, you know,

37:23

you could route. Like you could

37:25

almost replace like an MPLS tunnel

37:27

with a zero-tier tunnel Right because

37:29

then the zero-tier tunnel will punch

37:31

through anything it can create a

37:33

point-to-point connectivity And so all you

37:35

would need is like a zero-tier

37:37

interface on the customer side of

37:40

things and a zero-tier interface somewhere

37:42

else and you could immediately create

37:44

a point-to-point connectivity and so you

37:46

could almost see it almost like

37:48

as an MPLS type of tunnel

37:50

to to run a different kind

37:52

of service. You know, so you

37:54

could have internet going off one

37:56

direction. And then let's say you

37:59

wanted to offer, you know, full

38:01

BGP tables, phone or something, a

38:03

different service. Yeah. You could route

38:05

that through the zero tier portal.

38:07

So there is some differentiation you

38:09

could set up. Yeah. Yep. And

38:11

they have, you know, a lot

38:13

of like flow rules and stuff

38:16

like that that I haven't messed

38:18

with. But again, they have great

38:20

documentation and, you know, they, oh,

38:22

they do support IPB6. I think

38:24

I was going to ask. And

38:26

so you can have, you know.

38:28

You can set up your network

38:30

to be private, public, you know,

38:32

single sign on, you know, whatever

38:35

you'd like. And, you know, they

38:37

do bridging and kind of whatever

38:39

you need. And so that's one

38:41

of the reasons, you know, micro-tick

38:43

is extremely cost effective. The priced

38:45

performance is amazing. We've never had

38:47

any issues running micro-tick. And with

38:49

a zero tier on top of

38:51

it, the... Your imagination

38:54

is pretty much the limit on

38:56

on what you can do Something

38:58

you mentioned along the way is

39:01

that zero tears end-to-end encrypted. So

39:03

does that mean I have to

39:05

deal with certificate management that kind

39:08

of thing? Nope, they handle all

39:10

that You don't have to deal

39:12

with any of that. What got

39:15

to be something I got to

39:17

manage? Yeah, no, nope, nothing all

39:19

you need to do is install

39:22

the pack on micro tick and

39:24

even you know Windows or anything

39:26

like that all you have to

39:29

do is download it install it

39:31

connect to the network and that's

39:33

it it's they've made it super

39:35

simple and kind of bulletproof if

39:38

hold on a second guys actually

39:40

let me log into my router

39:42

really quick and I let me

39:45

actually see let me actually look

39:47

at the certificate side of things

39:49

Like if I can't see that

40:02

Okay, yeah, here I can see, so

40:05

if I go to, my, my, I

40:07

can look at the zero two instance,

40:09

it has an identity, and it has

40:11

a public identity. And so it's probably

40:14

just automatically generous, that's almost like your

40:16

certificate, like your private level certificate, and

40:18

it's generating that automatically is what it

40:21

is. But again, you don't have to

40:23

deal with that, Elijah. You're not managing

40:25

certificates or appreciate keys or any of

40:28

those things. Zero tears just dealing with

40:30

that for you. Yeah, it has when

40:32

you install the package it has an

40:34

identity and a public identity and it

40:37

handles all of it it generates all

40:39

those values automatically in my micro tick

40:41

router I just go to zero tier

40:44

I hit the plus button and I

40:46

name it whatever I want I tell

40:48

it the network ID and It does

40:51

what it wants you have to specifically

40:53

tell it to route global IP addresses

40:55

or public IP addresses. By default, it

40:57

will not route public IPs. So if

41:00

you want to route public IPs, you

41:02

got to watch out for that a

41:04

little bit. But yeah. Now, if we

41:07

think of the solution that you've built

41:09

here, we could describe it as basic

41:11

SDWAN or maybe DMVP is a better

41:13

analogy to what you've built. What would

41:16

you say the tradeoffs are with what

41:18

you've got versus some fancy full blown

41:20

SDWAN solution from a name brand vendor?

41:24

Well, that's a really good question. I

41:27

don't have a whole lot of experience

41:29

with some of the name brand vendors.

41:31

And so I don't know exact some

41:34

of the things you can get with

41:36

those guys. I'm sure they're great and

41:38

I'm sure they have a specific use

41:41

case. And, you know, if they have

41:43

a feature that you absolutely need, you

41:45

know, again, the right tool for the

41:48

job, but for us, the, The zero

41:50

tier SDWAN solution for for VPN or

41:52

again the SDWAN solution for like a

41:55

backup or whatever you need it works

41:57

really well and it's simple. and it's

41:59

easy and it's affordable and it works

42:02

good. And so again, I think, it's

42:04

really your imagination is the limit on

42:06

which you can do with it. Well

42:09

I also think it's tied to your

42:11

specific use case in that you're an

42:13

ISP where the thing that you need

42:16

more than anything else is connectivity. I

42:18

know with some full loan SDWans, I

42:20

mean SDWans kind of become a subset

42:23

of SACI mean it's really a different

42:25

use case. It's an enterprise and they

42:27

have security needs and they want on

42:30

demand routing and they want active active

42:32

links maybe and. And maybe they want

42:34

policy based routing that, or let's call

42:37

it app based routing where link by

42:39

link they can, depending on the characteristics

42:41

of the link in any given time,

42:43

they can pump an application across one

42:46

link or the other, depending on the

42:48

real time conditions of what that link

42:50

is. So that's a different animal. That's

42:53

not what you're delivering. That's not what

42:55

you're exactly doing here. You're like. Let's

42:57

keep Elk City online. How do we

43:00

do that in a transparent way that

43:02

works for them and is simple for

43:04

us to maintain? It doesn't cost a

43:07

fortune. That's a different answer. So yeah,

43:09

I really think it goes back to

43:11

what you were just saying. It's like,

43:14

it's a use case thing. Yeah, there's

43:16

probably some features there that, you know,

43:18

if you need them, you need them,

43:21

we don't. And so there you go.

43:23

Pretty much. If you need, you could

43:25

do, you know, with micro, between micro-take

43:28

and zero-tier, you could do that application-based

43:30

routing, you could do layer seven, firewall

43:32

rules, micro-rule, mango rules, you could do

43:35

all that, zero-tier has the flow controls,

43:37

but like you said, I don't need

43:39

them, I haven't used them, but if

43:42

you needed to, you could, you could

43:44

do all that kind of stuff with

43:46

it, but I never needed it. Elijah,

43:49

just I'm curious, thinking bigger as a

43:51

wireless internet service provider, what your take

43:53

is on satellite broadband in general, do

43:56

you see, obviously you're using it, you

43:58

know, as a backup. when your service

44:00

because of issues can't get there. But

44:03

do you see this as a competitor

44:05

down the line at some point? I

44:07

see all the tools in the toolbox

44:10

as being valuable. So I think fiber

44:12

has its place. I think fixed wireless

44:14

has its place. And I think satellite

44:17

has its place. I don't think satellite

44:19

will ever compete with fiber. And I

44:21

don't really ever see competing with a

44:24

good. fixed wireless solution. So I see

44:26

satellite being great to hit some of

44:28

the places that even we can't hit.

44:31

There's some places out here that are

44:33

just in the middle of nowhere and

44:35

they need connectivity and you know Starlink

44:38

and all the other slower satellite solutions

44:40

are a perfect solution what they need

44:42

and honestly about what they're only going

44:45

to get in some of these places.

44:47

But you know our fixed wireless solutions,

44:49

you know, there's some new 60 gigahertz

44:52

stuff coming out and You know we

44:54

can offer some of these 100 by

44:56

100 200 200 300 even some gig

44:59

speeds over wireless these days And so

45:01

I think with a good fixed wireless

45:03

solution, you're gonna kind of have a

45:06

tiered model where fiber is In our

45:08

area, you're going to have fiber, which

45:10

is really great, maybe in town, fiber

45:13

to the home. You're going to have

45:15

fixed wireless, going to be in this

45:17

intermediate between the town and, you know,

45:20

the mountain and some harder to get

45:22

two places, but still pretty good. And

45:24

then satellite be great farther out. So

45:27

I don't really, I see all the

45:29

technologies somehow merging together. I don't think

45:31

one is. for say, going to kill

45:34

any of the others. And I kind

45:36

of, that's kind of how I view

45:38

things. Okay, that makes sense. Yeah, it's

45:41

kind of what I was imagining with

45:43

Starlink. It's been interesting to watch Starling's

45:45

progress. They keep putting more birds up

45:48

in the sky, but then as more

45:50

people get onto the system, it does

45:52

seem to be affecting overall throughput. The

45:55

technology is incredible. but as a, you

45:57

know, placing what you're delivering, no, I

45:59

don't see it for that. It does

46:02

feel like it's, I know, I mean,

46:04

Starlike does have that package for ISPs

46:06

where it's a whole different kind of

46:09

a base station arrangement and so on,

46:11

but they. They want to, they want

46:13

you to fund building it. It's like,

46:16

I think it's 10 million just to

46:18

get in the door, something like that,

46:20

with what that is. But it's, but

46:23

it's delivering a huge amount of bandwidth,

46:25

you know, and an uplink to, to

46:27

the satellite constellation. So there is a

46:30

model for that, but again, making the

46:32

R.I work, that's, that's a tough one,

46:34

because it's a significant infrastructure investment to,

46:37

to, to get that star link, ISP

46:39

style base station up and off the

46:41

up and off the off the ground.

46:44

Well, Elijah, great conversation, man. Great conversation.

46:46

How do people contact you if they

46:48

have questions about what you built? Maybe

46:51

they want to get into zero tier.

46:53

They want to go, Elijah, can't figure

46:55

this out, help me out, man. Yeah,

46:58

if anybody wants to get into zero

47:00

tier micro-tech, ubiquity, whatever, you know, I'd

47:02

love to love to help out. You

47:05

can contact me on LinkedIn. You can

47:07

search for me, Elijah, Zeta. You can

47:09

send me an email, Elijah at Airbridge

47:12

Broadband.com. I'm typically not on the Twitter's

47:14

or the X's, it's called these days.

47:16

I'm not really there much, but I'm

47:19

on LinkedIn or if you can send

47:21

me an email. I'll add

47:23

to thanks for reaching out to us

47:25

to tell this story. This is a,

47:27

this is fun. Solutions like this are

47:29

great. And because of where I live

47:31

in rural New Hampshire, I love rural

47:33

areas and figuring out how to get

47:35

them connected is a puzzlement. It's funny

47:37

how mountains of all things really. complicate

47:39

life. They really make it difficult. This

47:41

is just infrastructure doesn't go up there

47:43

unless it's a tower and even that

47:45

it can be tenuous. So I love

47:47

what you're doing. And I want to

47:49

say thanks for your service almost. It's

47:51

almost like you're in there in deep,

47:53

you know, helping out people who need

47:55

help. So well, you know, we are

47:57

changing people's life. I will say, you

47:59

know, we've seen in this area a

48:01

ton of people move in being able

48:03

to work from home, you know, five,

48:05

six years ago, you couldn't work from

48:07

home in this area, and people can

48:09

do that these days. So I do

48:11

feel a little bit like we're changing

48:13

lives. Yeah, that's cool. Well, all right,

48:15

if you're still out there listening, thank

48:17

you very much for listening to heavy

48:19

networking today. Part of the packet pushers

48:21

podcast network, of course, find out all

48:24

about packet pushers at packet pushers.net. And

48:26

while you're there, we would appreciate it

48:28

if you'd sign up for a weekly

48:30

newsletter, Human Infrastructure, Drew and I, and

48:32

sometimes guest writers, serve up the very

48:34

best networking and IT engineering content that

48:36

we find on the internet and share

48:38

it with you. I was working on

48:40

this week's issue just before we recorded

48:42

this with Elijah. By the way, it's

48:44

not just links in there, but it's

48:46

also our own summaries and takes on

48:48

the post that we're reading, and we

48:50

also throw in nerdy memes of the

48:52

week that cracked us up and highlight

48:54

vendor news that we think you'll be

48:56

interested in. And of course, your contact

48:58

info is safe with us. We never

49:00

ever share your addresses with anyone, even

49:02

when they ask. And they do ask.

49:04

There is a lot of things going

49:06

on at Packa Pushes these days. We've

49:08

launched two new podcasts pretty recently, Technically

49:10

Leadership, with host Laura Santa Maria, all

49:12

about how to transition into management, be

49:14

an effective technical leader and related topics.

49:16

And then total network operations with host

49:18

Scott Rabon is focused on effective net-ops

49:20

because net-ops is not just working through

49:22

ticket after ticket in your bucket. Last,

49:24

but not least, remember that too much

49:26

networking would never be enough.